Modulate plant nitrogen uptake

[luyssaert]

If plants experience C-limitation they keep taking up nitrogen through their roots. This has been observed to result in very high N stocks in the labile and reserve pools.

[luyssaert]

Two changes were committed in r6869: (1) the labile and reserve nitrogen pools are now linked to each other. Nitrogen is assumed to freely move from one pool to the other whenever needed. This fix overcomes an earlier observed problem where the labile N pool was depleted but there was still plenty of N in the reserve pool. The plants would die or show strong signs of N-limitation despite the abundant presence of N in the reserve pool. (2) target n labile and reserve pools are calculated, if the target is reached, N-uptake from the roots is modulated.

[luyssaert]

Tests with r6869 showed improved LAI timeseries for most forest PFTs. Exceptions to this observation are PFT3 and PFT2. PFT3 was still problematic before these changes. Previous versions of the model nicely reflected the precipitation gradient over Africa in the LAI of PFT2. The additional N-feedback seems to partly break this pattern despite resulting in higher LAI than before.

The LAI timeseries of grassland and cropland PFTs became problematic after the implementation of the N-feedback. The calculation of this feedback uses the target n-pools which largely differs between trees and the other PFTs. This calculation of the target n-pools seems like a good place to start looking. We expect this problem is related to ticket #728.